Image Character Recognition Using Deep Convolutional Neural Network Learned from Different Languages

Jinfeng Bai,Zhineng Chen,Bailan Feng,Bo Xu

DOI: https://doi.org/10.1109/icip.2014.7025518

2014-01-01

Abstract:This paper proposes a shared-hidden-layer deep convolutional neural network (SHL-CNN) for image character recognition. In SHL-CNN, the hidden layers are made common across characters from different languages, performing a universal feature extraction process that aims at learning common character traits existed in different languages such as strokes, while the final softmax layer is made language dependent, trained based on characters from the destination language only. This paper is the first attempt to introduce the SHL-CNN framework to image character recognition. Under the SHL-CNN framework, we discuss several issues including architecture of the network, training of the network, from which a suitable SHL-CNN model for image character recognition is empirically learned. The effectiveness of the learned SHL-CNN is verified on both English and Chinese image character recognition tasks, showing that the SHL-CNN can reduce recognition errors by 16-30% relatively compared with models trained by characters of only one language using conventional CNN, and by 35.7% relatively compared with state-of-the-art methods. In addition, the shared hidden layers learned are also useful for unseen image character recognition tasks.